Saccharomyces cerevisiae actin-Escherichia coli lacZ gene fusions: Synthetic-oligonucleotide-mediated deletion of the 309 base pair intervening sequence in the actin gene

31

Gene. 22 (1983) 31-39 Elsevier

Saccharomyces cerevisiae actin-Escherichia coli 1acZ gene fusions: synthetic-oligonucleotide-mediated deletion of the 309 base pair intervening sequence in the actin gene (Recombinant exons; colony

DNA; oligonucleotide-directed in vitro mutagenesis; screening; /3-galactosidase; shuttle plasmids)

RNA splicing

and processing;

introns;

Garrett P. Larson, Keiichi Itakura, Hirataka Ito and John J. Rossi * Department of Molecular Genetics, City of Hope Research Institute, (213) 357-9711

1450 E. Duarte Road, Duarte, CA 91010 (U.S.A.)

Tel.

(Received October 6th. 1982) (Accepted December 22nd. 1982)

SUMMARY

Plasmids carrying gene fusions between the yeast (Saccharomyces initiation-defective Escherichia coli 1acZ (P-galactosidase) gene have P-galactosidase in such fusion plasmids depends on transcription after the RNA-splicing machinery has removed from the primary

actin gene and constructed. Expression

cerevisiae)

been

an of

of the actin gene, and is possible only RNA transcript the 309-bp intervening

sequence (IVS) interrupting the actin coding region. Mutants deleting the actin IVS were constructed via synthetic oligonucleotide-mediated in vitro mutagenesis of the actin-/3-galactosidase fusion plasmid. A 17-base synthetic oligonucleotide was used to generate a 309-bp deletion which precisely removed the actin IVS. A partial deletion mutant was also constructed in which 272-bp, starting at the 5’ end of the actin IVS, and including the 5’ splice junction signal, were deleted. Both the complete and partial IVS-deletion mutants were transformed into yeast hosts. However, the partial deletion resulted in a greater than 98% reduction

in P-galactosidase

/I-galactosidase

activity

activity.

as compared

The precise

deletion

with the parental

of the actin

fusion plasmid

IVS did not reduce

containing

the intact

the levels of IVS.

INTRODUCTION

l

To whom correspondence is to be sent.

Abbreviations:

(IIT, autonomously

replicating sequence; bp,

base pairs; CCC, covalently-closed circular;

EtBr, ethidium

bromide; IVS, intervening sequence (intron); kb, kilobase pairs; SD. see MAT. AND METH. (section b); SDS, sodium dodecyl sulfate; SET. see MAT. AND METH. (section k); SSC, 0.15 M NaCl, 0.015

M Nas’citrate, pH7.6; Xgal, 5-bromo-4-chloro-in-

doyl-p-D-galactoside; YEPD, see MAT. AND METH. (section b):

A, deletion.

0378-I 119/83/0000-OOW/$O3.00

Gene fusions serve as a powerful method for studying transcriptional and translational regulatory signals. In yeast, gene fusions have been constructed between cytochrome c (CYCI ) (Guarente and Ptashne, 1981), or URA3 (Rose et al., 1981), and an initiation-defective E. cofi IacZ gene segment. These fusions display the patterns of regulation normally seen for the yeast genes.

0 1983 Elsevier Science Publishers B.V.

32

We describe gene

and

cerevisiae pression peptide and

fusions

a defective shuttle

depends

plasmid

plasmid,

of the fused

removal

between

actin

m which functional

actin-j?-galactosidase

upon splicing was used

ex-

Mutational identification

actin-lacZ

as the substrate

in known

in a number

of the desired

DNA

fusion

Therefore,

mutant

synthetic

however,

can be difpheno-

oligodeoxyribonucleo-

tides (oligonucleotides) as mutational agents are useful since such directed mutants are very numerous and can be easily identified against a background of nonmutant DNAs, using the oligonucleotide as a specific hybridization probe (Wallace et al., 1980). Two systems are generally employed for oligothe singlenucleotide-mediated mutagenesis: stranded

DNA

MATERIALS

AND METHODS

for in

sequences

of ways;

ficult due to the lack of a readily detectable type.

gene product.

(a) Materials

changes

be effected

of the actin of the fused

poly-

vitro mutagenesis. can

tion, we show that a precise deletion IVS has no effect on the expression

of the actin mRNA

of the IVS. This

DNA

the yeast

IacZ gene in an E. coli-S.

phages,

such as M13. or single-

stranded plasmid DNA. The latter system generally requires fewer steps since it is not necessary to subclone the target DNA region into the phage; instead, single-stranded plasmid DNA is produced by first nicking the DNA, and then removing the nicked strand with exonuclease III. This procedure, using plasmid DNA, was used to delete a 14-bp IVS in a yeast suppressor tRNA gene (Wallace, 1980). The actin gene of the yeast S. cerevisiae contains a 309-bp IVS, which interrupts the amino acid coding region (Ng and Abelson, 1980; Gallwitz and Sures, 1980) but is excised from the primary transcript by a splicing mechanism. The IVS of the actin gene shares the same characteristics as other higher eukaryotes in that it starts with the dinucleotide 5’G-T3’, and ends with the dinucleotide 5’A-G3’, with a consensus sequence of six to eight nucleotides at both splice sites (Ng and Abelson, 1980; Gallwitz, 1982). In this report, we test the role of the yeast actin IVS in expression of an actin-IacZ gene fusion by making partial or complete deletion mutants of the actin IVS, using a synthetic oligonucleotide to produce deletions of 272 and 309 bp. Our results demonstrate the feasibility of large deletions using synthetic-DNA-mediated mutagenesis. In addi-

Restriction polynucleotide (large

endonucleases, kinase, and

fragment)

Laboratories.

were

T4 DNA

from

[ y- “‘P]ATP

DNA ligase. polymerase I

Bethesda

Research

was synthesized

by the

method of Walseth and Johnson (1979). or purchased commercially from ICN (crude > 5000 Ci/mmol). [a- 32P]dNTPs and exonuclease III were from New England Nuclear. Glusulase was from Endo Laboratories. (b) Strains and media E. coli strain MC1061 (araD139. A(ara. leu)7697, AlacX74, galUP, galK-. hsr-, h.sm+. strA) (Casadaban and Cohen, 1981) was supplied by M. Casadaban. formations

Selective medium for bacterial transwas L-broth (Miller, 1972) supple-

mented with 30 pg/ml S. cerevisiae strain

ampicillin. NNY (mata,

trpl,

gal?,

gallO, ura3-52, his3A-1), supplied by J. Wallis, was used for all yeast transformations. Liquid growth medium (YEPD) for yeast was 2% peptone, 2% yeast extract, and 1% glucose. Minimal plates (SD + Xgal) for yeast transformations sisted of 0.67% yeast nitrogen base without

media conamino

acids, 2% glucose, 2% agar, 20 pg/ml uracil. 20 pg/ml L-histidine, 40 pg/ml Xgal, and for buffering, 1 X M9 salts. M9 salts contain per liter: 6 g Na,HPO,, 3 g KH,PO,, 1 g NH,Cl, 0.5 g NaCl. (c) Bacterial and yeast transformations E. coli cells were transformed

by the Kushner

(1978) method. Yeast spheroplasts were prepared using glusulase, then transformed by the method of Beggs (1978) with only minor modifications. (d) Preparation of plasmid DNA and restriction enzyme analyses Plasmid

DNA

was isolated

grown in 0.5% Casamino

from

E. co/i cells

acids. 1 x M9 salts, 0.28

33

glucose,

1 mM MgSO,,

thymine. cleared

HCl, and

30 pg/ml

lysate procedure

was separated Bio-Gel

10 mM CaCl,,

from RNA

A-50m

carried

(Bio Rad).

CCC

pH

using

7.9,

mercaptoethanol,

6.6

plasmid

equilibrium 1969).

with restriction

out at 37°C

Tris . HCl,

using

a

ligation

1 unit T4 DNA

ligase in 10 ~1 of

buffer at 4°C for 18 h.

EndoR

buffer

MgCl,,

(i) Preparation of nicked plasmid pYAB1

DNA 200 pg of CCC pYAB1

centrifuga-

endonucleases

mM

on

were (10 mM

6 mM

2-

was nicked

using

the

procedure of Wallace et al., (1981) in 304 ~1 of Hin buffer and 300 ng/ml DNase I (Worthington). Nicked bromide Helinski,

and 60 mM NaCl).

nicked (e) Oiig~~xy~~nucleotide

using

et al., 1979). DNA

by chromatography

was purified by CsCl-EtBr tion (Clewell and Helinski, Digestions

ampicillin,

(Kahn

pGL007

1 pg/ml

plasmid

was purified

equilibrium

by CsCl-et~dium

centrifugation

1969). Approx.

(Clewell

and

70% of the DNA

was

by this procedure.

synthesis (i) In vitro mutagenesis

Two oligonucleotides, a 17-base actin 1% ““delete? 5’-AGCAACCTCAGAATCCA-3’ (174 IVS) and an unspliced CAGAATCCA-3’ the triester 1982).

actin probe 5’-GAACATAC(17-Act), were synthesized by

method

on a solid support

(Tan et al.,

5 pg (0.85 pmol) of nicked pYAB1 cubated with 35 units of exonuclease III of 50 mM Tris + HCl, pH 8.0, 10 mercaptoethanol, 5 mM MgCl, at 37’C min. The reaction was phenol extracted DNA was ethanol precipitated.

was inin 26 ~1 mM 2for 30 and the

(f) DNA sequencing

The exonuclease III treated DNA was resuspended in 6 ~1 of TE buffer (lo mM Tris . HCl,

The restriction-enzyme-digested DNAs were labeled at their 3’-termini with DNA poiymerase I (large fragment) and the appropriate [cw-32P]dNTP. DNA sequences were determined using the method

pH 8.0, 1 mM EDTA); 92 pmol of 5’ phosphorylated 17 base actin IVS deleter (17AIVS) was ad-

of Maxam

and Gilbert

(g) Cons~etion Approx. Hindlll-BglII

(1980).

of aetin-~-g~aetosidase 1.5 pmol fragment

fusions

of gel-purified 1.6-kb from pGJOl3 (previously

constructed in this laboratory) was joined to 0.35 pmol of ~~~dIII-~u~HI pXJOO3 (Rossi, J., Soberon, X., Marumoto, Y., McMahon, J. and Itakura, K., manuscript in preparation) after filling in the Bglll and BarnHI termini. The 20 ~1 ligation reaction buffer contained 66 mM Tris * HCl, pH 7.5, 6.6 mM MgC12, 0.4 mM ATP, 10 mM dithiothreitol, and 1 unit T4 DNA ligase. The reaction was carried out at i3”C for 6 h.

ATP and 170 pg/ml of each deoxynucleoside triphosphate were added (final volume 12 ~1). The reaction mixture was incubated at 12°C for 1 h, and used to transform

E. cofi strain

MC106 1.

(k) Colony screenings and ~ansformants Random screenings of transformants arising after overnight growth were carried out by blotting

replicating yeast

the cells to Whatman 541 filter paper, then amplifying the plasmid DNA on L-broth agar piates containing 250 pg/ml chloramphenicol for 16-20 h. The filters were then prepared for hybridization as described by Gergen et al. (1979). Prehybridizations and oligonucleotide hybridiza-

3 pmol of gel-purified 1.45-kb TRPI, fragment from pYRp7 (Carbon and 1980) was joined to pGL077 or

tions were carried out as described (Reyes et al., 1981), except SET was used in place of NET (1 x SET is 0.15 M NaCI, 30 mM Tris - HCI, pH 8.0, 1 mM EDTA), and hybridizations were carried out for 2-4 h. Both ohgonucleotide probes

(h) Preparation plasmids Approx. am, EcoRI Tschumper,

ded, and the mixture was heated in a boiling water bath for 3 min, then cooled to 4°C. To this mixture, 2 units T4 DNA ligase, 1 unit DNA polymerase I (large fragment), EndoR buffer, 5 mM

of autonomously

34

were

hybridized

at 46°C

and

the

filters

were

generated

by the desired

washed at 40°C in 6 x SSC. Restriction hybridizations were carried out as

fragment described

This in-phase

(Gergen

was per-

(Lac-)

et al., 1979). Autoradiography

formed

at -70°C

(with Kodak

film), using a DuPont screen and exposing (I) @Galactosidase

Lightning

XPR-I

Plus intensifying

the out-of-phase

was denoted

construction

pGL007.

(b) Cons~etion

of autonomously

replicating yeast

pksmids

for 2-24 h. assays

were centrifuged

while

end construction.

(Lac+) was designated

or XAR-5

A 1.45-kb EcoRI fragment, containing the yeast TRPl gene and a yeast origin of replication, the

Yeast cells from 0.5 to 5.0 ml of minimal culture

as pGLO77,

blunt

construction

media

at 3000 x g for 5 min and

resuspended in 1.O ml of Z buffer (Miller, 1972). Two drops of chloroform and 1 drop 0.1% SDS were added and the tube vortexed for 10 s. The enzymatic assays were carried out as described (Miller, 1972).

ars gene, was inserted

into the EcoRI

pGL077

(Fig. 1). The new plasmids,

and pGL007

sites of both

which contain the TRPI and ars genes upstream from the actin promoter region, were designated pYAB1 and pYAB0, respectively. When transformed into a yeast host, only pYAB1 gave Lac’ colonies. (c) In vitro mutagenesis Nicked plasmid pared as described

RESULTS

(a) Construction

of yeast actin-b-galactosidase

fu-

sions As shown in Fig. I, a 1.6-kb HindHI-BglII fragment (BgZII site filled in by large fragment DNA polymerase I) from pYACT-I, containing the upstream transcriptional regulatory region for the yeast actin gene, the entire 309-bp IVS and the first 252 nucleotides of the actin coding sequence, was ligated with the large HindIII-BarnHI fragment (BarnHI site filled in by DNA polymerase I) of pXJOO3. The plasmid pXJOO3 is a derivative of pMC1403 (Casadaban et al., 1980) carrying all but the first 8.1/3 codons of IucZ. The resulting ligation mixture was used to transform E. co/i. The transformants, plated on SD + Xgal, were screened with a 17-base actin splice junction probe (17-Act), which spans nine bases on the 5’ side of the splice junction and the first eight bases of the IVS. Both Lac+ and Lac- transformants were obtained. DNA sequence analysis of the actin+galactosidase gene fusion points in plasmids from both types of colonies, revealed that the Lac- phenotype was due to the ligation of the BgiII and BarnHI cohesive termini, resulting in actin codons being out of phase with la&. The Lac’ phenotype was the result of an in-phase translational reading frame,

DNA from pYAB1 was prein MATERIALS AND METHODS.

section i, and exonuclease III was used to degrade the nicked strand, producing single-stranded, circular templates for oligonucleotide-primed polymerization. The 5’-phosphorytated oligonucleotide, complementary to 9 bp immediately 5’ to the IVS and 8 bp on the 3’ side (Fig. 2A), was annealed with the single-stranded circles. Primed DNA synthesis was initiated with DNA polymerase 1 (large fragment) and all four deoxynucleoside triphosphates in the presence of T4 DNA hgase. The resultant heteroduplexes were used to transform E. cofi strain

MC 106 1. Approx. 20 000 colonies were screened (Fig. 3B) using the actin IVS “delete? as a highly specific hybridization probe. The initial screening

revealed

five positive

hybridization

sig-

nals. Upon a second screening, four of these five colonies reproduced positive signals (Fig. 3C). Plasmid DNA was prepared from these four mutants and analyzed by HrpaII restriction digestion (Fig. 3D). Mutant 3 (M3) produced a band corresponding to a fragment size of 466 bp. the size predicted if the IVS was precisely deleted. Mutants M2 and M4 are altered outside the target region (indicated by arrows), and have not been characterized further. Ml appears to be a heterogeneous mixture of both correctly nonmutant plasmid DNAs.

deleted

and

35

IVS EcoRI

HindIII vt”,y,‘,“,

I

TrpI

)

in

EcoRI

Ars

I

PstI

Ori

1

EcoRI

EcoRI

/

Ori

PstI

Fig. 1. Construction

of yeast actin+-galactosidase

with the large HindlII-BornHI

fragment

fusion plasmids.

cohesive

Hind111 termini

and (ii) the flush-ended

plasmids

were designated

pGLOO7 or pGL077

was inserted

into the actin-lacZ

from the actin initiator promoter.

AUG

141 bp upstream

(see RESULTS).

fusion vector as shown. into IacZ is depicted

(d) Characterization actin TVS

(pYAB1).

Transcription

from pYACT-I

(from pXJOO3) termini

1.4-kb EcoRI

shuttle

fragment

pXJOO3. This was accomplished

and BumHI

(2) A purified

The resultant

of the first actin ATG (Gallwitz,

plasmid

restriction

of the actin-lncZ

isolated

translational

fusion message

(i) the

as shown. The resultant

fragment

with an in-phase

was joined

by joining

initiates

from pYrp7 reading

frame

from the actin

1982).

were obtained. of a partial deletion of the

Under the same conditions, delete the actin IVS resulted

plasmid

(filled in) Bg/II (actin)

DNA sequence analysis of M3 confirmed that an exact deletion had been made (Fig. 3E). This plasmid was designated as pAAIVS. When pAAIVS was transformed into a yeast host, only Lac+ transformants

(1) The 1.6-kb Hin dIII-BglII

of the E. coli lac.Z-containing

event. Of approx. 7500 colonies screened with the synthetic oligonucleotide, only one positive colony was isolated. The DNA sequence at the IVS-exon junction was determined from the isolated plasmid DNA. The data indicated that 272 bp at the 5’ end of the IVS were deleted, while 37 bp at the 3’ end of the IVS were left intact; additionally, 3 bp of the oligonucleotide 2B). This plasmid,

another attempt to in an unexpected

sequence were inserted designated pSAIVS,

(Fig. when

transformed into a yeast host, gave Lac- colonies. However, direct /3-galactosidase measurements from yeast cells harboring this mutant plasmid

36

@

5’ UPSTREAM

TAACAATGGATTCTi%%GTTGCTGCT

#3-wL

w

3’.ACCTAAGACTCCAACGA-5’ 272bP

@

1”s

TbCfN.

a

5’ UPSTR.EAM

nct,n

TAACAATGGATTCTGTTGCTV3’.ACCTAAGA

cm-

IYS

p

GAL

AGGTTGCTGCTW

5’

1

0

MET

S’UPSTRELM

_

32t.p

PCfl”

>-GAL ,_

IV.5

------i~AGGTTGCTGCT

TAACAAGG-TCTGAGGTTGCT

T

, OUT OF PHKZ TRINSLITlON INTO p-&c )

Fig. 2. Translational

reading

IVS leads to an in-phase actin exon boundaries (17AIVS)

to pYAB1

out-of-phase

frames

reading

single-stranded

deletion

template.

after transformation section

METHODS

mutants

the deletion

DNA

during

mutagenesis A synthetic

of 5’ partial

of the actin IVS. (A) The precise deletion

of the 17-base “IVS-delete?

deletion

in vitro mutagenesis.

mutant

showing

base pairing

(C) The 5’ partial

deletion

were initially

k. (C) Initial positives demonstrated

to delete the actin IVS. (A) Nicked pYAB1 was digested

oligonucleotide screened

that mutants

bases represent

(17AIVS)

was used to prime

DNA

with the 17A IVS as a specific hybridization

were rescreened

to confirm

Ml and M3 contain

of the actin IVS. (E) DNA sequence

The underlined

deletions

Pairing

of ths actm

probe (lower sequence) of the I7-base of the actin

with the

actin IVS deleter IVS results

in an

into @-galactosidase.

Fig. 3. Oligonucleotide-directed arising

and (B) partial

into P-galactosidase.

is also shown. (B) Structure

translation

a single stranded

of (A) complete

frame

analysis

the left exon boundary

a positive

a restriction

hybridization fragment

(arrow)

in the target region of M3 confirmed

repair

with exonuclease synthesis

III generating

in vitro. (B) Colonies

probe as described

in

MATERIALS

signal. (D) Hpn II digestion corresponding

to the size expected

that an exact deletion

(G) fused to the first base of the right exon boundary

(A).

AX;D

of the four for

had been made.

37

TABLE

addition,

I

/3-Galactosidase ing various

activity

plasmid

of the S. cereursiae

NNY host harbor-

constructions P-Galactosidase activity

IVS

mutant,

(units)

interesting,

but unexpected

region

which

independent

has

been

was generated

in Fig. 3, was

pYAB1

236

k95

pA3 IVS

223

k85

of the 5’ end of the actin IVS, but leaving

0.6+

p5’A IVS

measurements

MATERIALS

and

AND

standard

sented

were derived

of hosts bearing measurements

hosts

or pAAIVS,

observed

greater

(usually

p5’.?t IVS-bearing

less than are

hosts

0.01 units verified

activities

were

is well above under

by

hosts vs. pYRp7-bearing

plates. The ~5’3 IVS-containing several days incubation

cycle. In no case

pYAB1 and pAAIVS-

15% when

ground

de-

colonies

the

phenotype

of

hosts on SD + Xgal turn pale blue after colonies

do not turn blue at all.

of P-galactosidase

by a pairing

of the

five bases at the 5’ end of the 17-mer to

an exact complement

at positions

247-252 (Ng and

pected configuration was apparently stable enough to have primed DNA polymerase I mediated re-

activity

The resultant sequence

plasmid

5’ splice junction

has lost the consensus signal, but maintained

intact 37 bp in the 3’ splice junction region. This construction had less than 2% of the ,&galactosidase activity found in the parental pYAB1 containing hosts (Table I). suggesting an alteration of splicing generated by this deletion. The newly created sequence (Fig. 2C), if translated from an “ unspliced” mRNA, beginning with the first actin AUG, would not be translated in the proper reading frame for /Lgalactosidase expression. This mutant,

along

with

another

of

our

fusions

(pYABO), in which the actin translational reading frame is out-of-phase with the /ucZ coding se-

DISCUSSION

We have

at the 5’ end of the IVS (see Fig. 2B).

These results are best explained terminal

as well as a 3-bp

pair synthesis.

back-

the conditions

while the pYRp7-containing

resulted in low levels (Table I).

272 bp 37 bp at

Abelson, 1980) in the actin IVS (Fig. 2B). At the 3’ end of the 17-mer, 9 bp paired precisely with the target sequence at the 5’ actin exon. This unex-

measurements

The low level of P-galactosidase

in the p5’AIVSbearing results

pre-

hosts. The measurements

between

than

simultaneously.

These

activities

and five independent

of the yeast growth

activity used).

(b). The mean values

from seven independent

of pS’AIVS-bearing

was the difference termined

section

out as described

of &galactosidase

pYAB1

were taken at all phases bearing

insertion

were carried

METHODS,

deviations

encompassing

the 3’ end of the IVS intact,

none detectable

d &Galactosidase in

I.2

a deletion

This

in an experiment

of the one summarized

to contain

in the

characterized.

found

PYRP7

muta-

tions were also generated (Fig. 3D). One mutant with an unplanned alteration actin

Plasmid

other

constructed

fusions

between

the S.

cerevisiae actin and E. coli 1acZ genes in shuttle plasmids capable of autonomous replication in either E. coli or S. cerevisiae. Expression of the actir+galactosidase fusion sequences in a yeast host was obtained (Table I). One of our goals in constructing the actin-lacZ gene fusions was to generate a substrate for synthetic DNA-mediated in vitro mutagenesis of the actin intervening sequence. The methodology for carrying out such mutagenesis on plasmid DNAs has been worked out by Wallace et al. (1980; 1981). We have extended the potential of this methodology by utilizing a 17-base oligonucleotide the actin 309-bp intervening

to precisely delete sequence (Fig. 3). In

quence, provides evidence actin-lacZ fusion message actin AUG,

that translation of the initiates with the first

and hence requires

correct

splicing

of

the actin IVS in pYAB1. This interpretation is strengthened in light of experiments reported by Gallwitz (1982). Gallwitz has constructed, in vitro, an actin gene deletion which extends from the 5’ splice 2 to actin Since

signal to the middle of the actin IVS (bases 164 of the IVS). In this mutant, unspliced mRNA is accumulated but not translated. our deletion plasmid p5’AIVS is very similar

to this IVS mutant (Fig. 2B, C), it can be argued that none of the in-frame, actin AUG codons downstream from the IVS can function as translation starts in yeast. Therefore, the translation initiation site in both pYAB1 and pAAlVS must

38

be at the normal

actin site upstream

our knowledge then, the product first such construction involving

of the IVS. To

of pYAB1 is the an mRNA splic-

ing event to generate a functional, fused, eukaryotic-prokaryotic protein-coding sequence. Complete

deletion

IucZ fusion

duced

IVS (p5’AIVS)

expression

demonstrated did

not

splice

while the partial

affect donor

(Table

resulted

actin GT

within

gene signal

expression described

partially,

With regard genesis using

or precisely

deleted

to the efficiency

P-galactosidase

activity

re-

oligonucleotide other sequences oligonucleotide

for

annealing

as

detection

of

in yeast cells. This study

also IVS

Cancer

unless above,

the

Research

Hope Research

Center

(CA16434)

at the City of

Institute.

was

REFERENCES Beggs. J.D.:

Transformation

plasmid.

IVS. muta-

mismatched

primers to the templates, allowing with incomplete homologies to the to form hybrids as well. In future

work, the frequency of occurrence of these nontarget area mutants might be minimized by the use of longer oligonucleotide primers and hence, higher stringency annealing conditions. The low efficiency of mutagenesis

pMC1034.

regarding

(1982)

of deletion

the

I. M. Berman.

the actin

unplanned mutants. These priming events can be best explained by the conditions of lowered strinused

pYACT

for supplying

of

a relatively short, synthetic oligonucleotide, the frequency of successful events was low (approx. 0.01%). Even so, several thousand colonies could be screened by hybridization with the synthetic 17-mer, which is a highly specific probe. It should be pointed out that priming at sequences other than those of the target area do occur, resulting sometimes in the generation of

gency

M. Casadaban

information

mutated. In these studies, unspliced actin transcripts accumulated in the cell. We are currently examining levels of RNA from cells transformed with our fusion constructions containing either an intact,

tance, R. Ng for supplying and

was supported by USPHS grants GM301 34 (J.J.R.) and GM30395 (K.I.). K.I. is a member of the

in greatly

I). Gallwitz

that deletions

deletion

The authors wish to thank M.J. Rossi. P.R. Reeve, D. Fraser and C.J. Adams for their assis-

well as useful

did not have any effect on expression

of the fused protein, the actin

of the actin IVS in the actin-

ACKNOWLEDGEMENTS

Nature

Casadaban,

M.J. and Cohen.

signals

by DNA

ies, as compared with results previously published (Wallace et al., 1980, 1981), can be ascribed in part to the lowered probability of looping out a 309-bp sequence and forming stable hybrids with a 17-base oligonucleotide. Nevertheless, our results clearly demonstrate the feasibility of precisely deleting large sequences with relatively short oligonucleotide primers. We have also demonstrated that a simultaneous deletion and insertion can be generated using synthetic oligonucleotide mediated mutagenesis (Fig. 2B).

by a replicating

hybrid

fusion

S.N.:

Analysis

and cloning

of gene control

in Escherichia

cd.

J.

Mol. Biol. 138 (1980) 179-207. Casadaban,

M.J.,

fusions

Chow

that join

segment

J. and

Cohen.

to amino-terminal

cloning

S.N.:

an enzymatically

of translational

active

fragments

In vitro

gem

p-galactosidase

of exogenous

pro-

vectorsfor the detection

teins: Escherichia coli plasmid

initiation

signals.

and

J. Bacterial.

143

(1980) 971-980. Clewell,

D.B. and Helinski,

D.R.:

Supercoiled

protein

complex

in Escherichra

duced

conversion

to an open

Natl. Acad. Gallwitz.

complete

Sums.

circular

sequence

DNA

DNAand

in-

form.

Proc.

of a split yeast

gene:

1159- 1166.

I.: Structure

nucleotide

circular

coii: purification

Sci. USA 62 (1969:

D. and

of the actin gene from Sac-

charom)?ces cereuisiae. Proc. Natl. Acad. Sci. USA 77 (1980) 25462550. Gallwitz.

D.: Construction

mutant

of a yeast actin gene intron deletion

that is defective

lation

of precursor

in splicing and leads to the accumu-

RNA

in transformed

yeast

cell. Proc.

Natl. Acad. Sci. USA 79 (1982) 3493-3497. Gergen.

J.P., Stem, R.H. and Wensink,

permanent

in these stud-

of yeast

275 (1978) IO&l08

collections

P.C.: Filter replica and

of recombinant

DNA plasmids.

Nucl.

Acids Res. 7 (1979) 2115-2136. Guarente,

L. and Ptashne.

to the cytochrome

M.: Fusion

of Escherichia cob IarZ

c gene of Soccharomyces

cerewisiae. Prcxz.

Natl. Acad. Sci. USA 78 (1981) 2199-2203. Kahn,

M.. Kolter,

Remaut. derived (Ed.),

E. and

Helinski.

from plasmids Methods

New York, Kushner,

R., Thomas.

S.R.:

C., Figurski. D.R.:

ColEl.

17-23.

R..

vehicles

Vol. 68 Academic

Press,

1979, pp. 268-280. An improved

S. Nicosia

North-Holland

D.. Meyer. cloning

F, R6K. and RK2. in R. Wu

in Enzymology,

method

Escherichia coli with ColEl-derived and

Plasmid

(Eds.), Biomedical

Genetic Press.

for transformation plasmids.

of

in H.W. Bayer

Engineering.

Elsetier/

Amsterdam,

1978.

pp.

39

A.M. and Gilbert,

Maxam,

with base-specific

W.: Sequencing

chemical

cleavages.

end-labeled

DNA

Methods

of Enzymoi-

Genetics,

Cold Spring

Harbor

Experiments Laboratory.

352-355,

in Molecular Cold

Spring

Harbor,

NY,

1972, pp.

J.: Isolation

actin from SaccharomJws

and sequence

cereorsiae.

of the gene for

Proc. Natl. Acad.

Sci.

A., Johnson.

C.. Itakura, H-2Kb-related genetics

M., Ito. H., Ike, Y., Morin,

Wallace,

molecule

by

R.B.:

Identification

molecular

cloning.

of an Immuno-

to fi-galactosidase in yeast.

M.J. and Botstein, in Escherichia Proc.

Natl.

Carbon,

containing

gene. Gene

J.: Sequence

a chromosomal

10 (1980)

R.B.. Johnson,

J.: Directed

RNA intervening

sequence.

R.B.,

Itakura,

K.:

Schold,

P-globin

Solid-phase

synthesis

of a yeast

replicator

Symp. DNA

and the TRPl

specific point mutations

M., Itakura,

of a yeast

transfer

Science 209 (1980) 1396- 1400.

M., Johnson,

gene:

S., Schold,

deletion

OIigonucleotide-directed a general

M.J.,

Dembek,

P. and

mutagenesis

of the

method

for producing

in cloned DNA. Nucl. Acids Res. 9

D.: Yeast genes fused

cdi can be expressed

Acad.

Sci. USA

nor-

78 (1981)

Walseth,

T.F. and Johnson,

R.A.:

T.. Dugaiczyk,

of polynucleotides,

A. and Itakura,

K.:

VII. The phos-

The enzymatic

of iY- 32P] nucleoside triphosphates, cyclic[.“P]GMP. Biochim. Biophys.

2460-2464. Tan. Z-K., Ikuta. S.. Huang.

Harbor

157-166. P.F., Tanaka,

K. and Abelson,

human

Cold Spring

(1981) 3647-3656.

14 (1981) 383-392.

Rose. M.. Casadaban, mally

M.J., &hold,

K. and

G. and

Wallace,

USA 77 (1980) 3912-3916.

solid phase method.

Biol. 47 (1983) in press.

fragment Wallace,

433.

Ng. R. and Abelson,

Reyes,

Quant. Tschumper,

ogy 65 (1980) 499-560. Miller. J.H.:

photriester

Communicated

by A.D. Riggs.

cyclic

preparation

[ 32P)AMP, and

Acta 567 (1979) 11-31.